ecommerce-ozon-product-trend
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes external product data from Ozon via a proxy. This creates a surface where content embedded in marketplace data could potentially influence the agent's behavior.\n
- Ingestion points: Marketplace data is ingested via the API response handled in
scripts/ozon_product_trend.py.\n - Boundary markers: While
SKILL.mdincludes display rules, it lacks explicit delimiters or instructions for the agent to ignore instructions potentially embedded in the data.\n - Capability inventory: The skill includes script execution (
scripts/ozon_product_trend.py) with capabilities for authenticated network requests and local file writes to the project directory.\n - Sanitization: The script does not sanitize API response values before providing them to the agent.
Audit Metadata